According to TAUS, we are headed toward the Age of Convergence, where translation memories, machine translation, leveraging tools, and search tools will converge into a cohesive user experience across devices. With so much evolving technology, what is the future of lifelong linguists and transcreators like me?
At industry venues as well as within the localization industry as a whole, people are talking about fully-automated localization workflows and bots that translate better than professional linguists. Unmanned translation systems and robot writers are breaching the threshold from science fiction to science fact. Everybody knows about free machine translation tools such as Google Translate. It’s quick, convenient, and provides excellent results depending on the language pair.
I’ve used its Korean to Japanese capabilities as a gisting tool and was amazed at the results. So far so good, but how about other pairs, like Japanese to English? In a bit of good news for Japanese to English translators, I’ve found the target text to be quite incoherent and about as gratifying as a conversation with a thoroughly-inebriated person.
So the quality isn’t there yet for a number of language pairs. But machine translation (MT) is on fire — not because of its quality, but its speed. Companies need to reach out faster to more people, pushing highly perishable content. The short lifespan and high frequency means time is of the essence, not quality. MT is satisfying this need, and companies are clamoring for their own optimized MT solutions. These may or may not include a human post-editing element.
So what about the future?
Uh oh. In this new worldview, do I see a linguistic future of post-editing ad infinitum, painting a thin veneer of humanity on machine-spewed strings while contending with inhumane turnaround times? In fact, with all the CAT tools and TMs we work with today, aren’t we linguistic cyborgs already? And in a day and age when AIs are defeating our grand masters at brainy games like chess and Go, isn’t it a matter of time before machines prevail and human translation becomes a thing of the past?
AlphaGo vs. South Korean professional Go player Lee Sedol: the final result. Credit: Google
Well, for one thing, we love new technology, so we will fixate on the rising role of MT. But on the other end of the spectrum there is still strong and growing demand for high-quality translation. When I asked Moravia’s CEO Tomas Kratochvil about this when he visited the Moravia Tokyo office en route to Beijing, his reply was “when companies need strong messaging for their brands, they require the highest quality transcreation. So the market will always exist for talented, knowledgeable, enthusiastic human translators. MT has its function, but it’s in the time-sensitive area that’s less about quality.”
In other words, the future will see more specialization in the linguistic services sector, that’s for sure, and linguists will be busy as ever in evolving roles as post-editor, translator, transcreator, reviewer, and more. Human-machine collaboration will continue at a range of levels and in a variety of new ways as we seek to streamline the process and offer more value to our customers.
So never fear, there will be plenty for an aspiring linguist to do as an in-house translator, working at an LSP, or freelancing. Especially if that linguist is open to embracing how his/her contribution will change in the light of technology. If you want to translate and expand your horizons on a global scale, the opportunities are and will be there. Luckily, for now at least, algorithms don’t understand colloquialisms, and as “quatre-vingt-dix” shows us, language is not a logical game like chess. Language is as human as we are.